Click to watchMTN Group has pledged to support the development of African language datasets, responding to a call by Nigeria’s Minister of Communications, Innovation and Digital Economy, Dr Bosun Tijani, for stronger public and private investment in AI research. This decision was made during The Y’ello Chair Vodcast at the United Nations General Assembly (UNGA) in New York.
The move is significant in a continent where more than 2,000 languages remain poorly represented in AI systems. Without structured datasets, Africa’s 1.5 billion people risk being excluded from AI-driven services in education, health, commerce, and governance.
Building Africa’s language models
The pledge builds on the Nigerian Atlas for Languages and AI at Scale (N-ATLAS). The project was launched earlier this year through a partnership between the Nigerian government and Awarri Technologies. N-ATLAS is an open-source multilingual model that digitises and preserves Nigeria’s 500-plus languages. Its framework is designed to be adapted by other African countries.
MTN Group President and CEO, Ralph Mupita, said the work is central to ensuring Africans are not left behind in the global technology shift. “We have to avoid the risk of Africans being a digital underclass. The outcomes we want are that people are digitally included, economically included and that they have dignity,” he said.
Related Article: MTN in talks with US and European firms to bring AI data centers to Africa
Click to viewBeyond cultural preservation, MTN’s support also points to new opportunities in Africa’s digital economy. By backing language datasets, the company is positioning itself as more than a telecom operator. It wants to be seen as a technology enabler in the continent’s AI space. Other global firms, such as Google and Meta, have also experimented with African language models, though usually with limited commercial scope.
MTN’s move highlights the growing interest of the private sector. Still, it is not the first initiative to tackle the language gap in African AI. Masakhane, a grassroots research community of machine learning experts, has been building translation models for African languages since 2019. Mozilla’s Common Voice project has crowdsourced speech datasets in languages such as Kiswahili, Luganda and Wolof. South Africa’s Lelapa AI is developing natural language processing tools for under-resourced African languages.
By joining this push, MTN brings corporate scale and resources to an effort long driven by academics and open-source communities. The outcome will determine whether African businesses can develop AI solutions grounded in local realities or continue relying on global systems that overlook them.
Click to watch
